Displaying data and separating 'U', 'O' and 'T' meteor trail observations

Separating U, O and T trail observations

In order to obtain an approximate profile of meteor activity associated with known major meteor streams the following will be assumed :

• Data will be filtered statistically to remove any observations having a duration (d) =or> 720 cycles (approx 120 secs). The base data thus filtered will be considered to contain all U+O meteor trails
• The base U+O data will be filtered to remove any observations having a duration (d) < 18 cycles . The data thus filtered will be considered to contain no U or T observations and designated O-T meteor trails. " We have, however examined a sample of such records and then by inference devised a method of statistical separation that is probably 99% accurate. Based on sample echoes, it was found that for durations longer than 3 secs almost no underdense decays were seen". DD Meisel, JE Richardson Planetary and Space Science 47 (1999). Similar iterative analysis of reflection events observed by the LVST confirms the relevance of this assumption for the LVST equipment and set up.
• The base U+O data will be filtered to remove any observations having a duration (d) =>18 cycles. The data thus filtered will be considered to contain no O observations and designated U+T meteor trails.
• There is no easy statistical way to remove T trails from the U set and add them to the O set in which they really belong. It is however possible to draw some inferences regarding T trails by examining other aspects of the data collected. In particular data proportional to the energy of the reflection may provide an effective filter in removing U trail reflection observations from the O+U set to create an O+T sub set.